Issue 40 Unmanned Systems Technology October/November 2021 ANYbotics ANYmal D l AI systems focus l Aquatic Drones Phoenix 5 l Space vehicles insight l Sky Eye Rapier X-25 l FlyingBasket FB3 l GCS focus l AUVSI Xponential 2021

8 Platform one Researchers in Greece have developed an integrated system for farms that they call the Autonomous Robot for Grape (ARG) harvesting (writes Nick Flaherty). The system, developed at the International Hellenic University in Thessaloniki, consists of three interdependent units: an aerial unit, a remote control (RC) unit and the ARG ground unit. The modular design means the system can be applied to a range of different crops and/or orchards. The ARG is designed and built to carry out three viticultural operations – harvest, green harvest and defoliation – which have a major effect on the quality of the grapes. They are considered to have a considerable effect on wine quality, so farmers are very particular about the harvesting process. The aerial unit, based on a SkyHawk UAV with a Samsung RGB camera, provides images of the vineyards. The RC unit uses the images to build the vineyard maps and define all the possible navigation paths. The UAV also allows an operator to design a path plan for the ARG to follow. The user can select between the different viticulture operations that can be personalised and sent to the ARG. The ARG navigates along a vineyard’s corridors, collects sensory data displayed to the RC unit and performs the selected viticultural operation in areas specified by the operator. The ARG uses a four-wheel Summit XL HL from Robotnik as the basic platform. This can carry a payload of up to 65 kg, which is enough for the manipulator together with all the embedded systems, and can navigate on uneven terrain with safety. An ORBBEC Astra 3D camera and Velodyne VLP-16 Lidar with a FLIR/ Teledyne A65sc thermal camera are used for positioning. A u-blox EVK-7P GPS and Pixhawk IMU provide positioning data to the central controller based on two Nvidia Jetson TX2 modules. This is powered by a 50 Ah 24 V lithium iron phosphate battery pack that has a lower energy density than lithium-ion but greater stability and safety. A ZED Mini 3D camera mounted on the robotic arm is the main visual sensor for the system. This provides a streaming video sequence that can be monitored from the RC unit for grape cluster and leaves detection, grape stem detection, estimating the ripeness of the grapes and identifying any defects. This is handled by an AI inference framework in the RC unit. Multi-vehicle systems Robot’s pick of the crop October/November 2021 | Unmanned Systems Technology The autonomous grape-picker can perform different operator-specified viticulture operations